https://aiweekly.co/alerts/gpt-6-astras-hidden-reasoning-extracted-via-api-tool-trick AI Weekly en 2026-09-24T15:54:49+00:00 GPT-6 Astra's hidden reasoning extracted via API tool trick https://aiweekly.co/alerts/speakermem-r1-posts-gains-on-four-multi-party-memory-benchmarks AI Weekly en 2026-09-24T15:54:49+00:00 SpeakerMem-R1 posts gains on four multi-party memory benchmarks https://aiweekly.co/alerts/trump-ally-memo-names-amodei-as-face-of-ai-doomerism AI Weekly en 2026-09-24T15:39:03+00:00 Trump-Ally Memo Names Amodei as Face of AI 'Doomerism' https://aiweekly.co/alerts/meta-pulls-dutch-satirists-ray-ban-glasses-video-over-bullying AI Weekly en 2026-09-24T15:39:03+00:00 Meta Pulls Dutch Satirist's Ray-Ban Glasses Video Over 'Bullying' https://aiweekly.co/alerts/google-eye-ai-scales-from-one-hospital-to-a-million-patients AI Weekly en 2026-09-24T15:12:01+00:00 Google eye-AI scales from one hospital to a million patients https://aiweekly.co/alerts/oracle-sends-force-majeure-notice-on-blue-owls-project-jupiter AI Weekly en 2026-09-24T13:57:21+00:00 Oracle Sends Force Majeure Notice on Blue Owl's Project Jupiter https://aiweekly.co/alerts/google-openai-anthropic-court-sriram-krishnan-for-ai-safety-body AI Weekly en 2026-09-24T13:57:21+00:00 Google, OpenAI, Anthropic Court Sriram Krishnan for AI Safety Body https://aiweekly.co/alerts/ssrn-pulls-257-papers-by-u-chicagos-nicholas-polson AI Weekly en 2026-09-24T12:38:03+00:00 SSRN pulls 257 papers by U. Chicago's Nicholas Polson https://aiweekly.co/alerts/nature-examines-ai-extinction-risk-after-coxon-resignation AI Weekly en 2026-09-24T12:12:02+00:00 Nature examines AI extinction risk after Coxon resignation https://aiweekly.co/alerts/nature-dissects-ai-extinction-claims-after-anthropic-exit AI Weekly en 2026-09-24T12:12:02+00:00 Nature dissects AI extinction claims after Anthropic exit https://aiweekly.co/alerts/amie-team-prospective-trials-not-benchmarks-earn-ai-trust AI Weekly en 2026-09-24T12:12:02+00:00 AMIE team: prospective trials, not benchmarks, earn AI trust https://aiweekly.co/alerts/amie-team-benchmarks-alone-cant-earn-medical-ais-trust AI Weekly en 2026-09-24T12:12:02+00:00 AMIE team: benchmarks alone can't earn medical AI's trust https://aiweekly.co/alerts/on-prem-medical-ai-agent-hits-90-accuracy-in-nature-paper AI Weekly en 2026-09-24T12:12:02+00:00 On-prem medical AI agent hits 90% accuracy in Nature paper https://aiweekly.co/alerts/amie-team-calls-prospective-medical-ai-trials-non-negotiable AI Weekly en 2026-09-24T12:12:02+00:00 AMIE team calls prospective medical AI trials non-negotiable https://aiweekly.co/alerts/nature-medicine-on-prem-clinical-ai-hits-90-on-mimic-iv AI Weekly en 2026-09-24T12:12:02+00:00 Nature Medicine: on-prem clinical AI hits 90% on MIMIC-IV https://aiweekly.co/alerts/amie-authors-medical-ai-trust-requires-prospective-trials AI Weekly en 2026-09-24T12:12:02+00:00 AMIE authors: medical AI trust requires prospective trials https://aiweekly.co/alerts/deepseek-revenue-hits-1b-run-rate-eyes-75b-shanghai-raise AI Weekly en 2026-09-24T10:58:19+00:00 DeepSeek Revenue Hits $1B Run Rate, Eyes $7.5B Shanghai Raise https://aiweekly.co/alerts/paper2agent-converts-research-papers-into-runnable-ai-agents AI Weekly en 2026-09-24T09:12:01+00:00 Paper2Agent converts research papers into runnable AI agents https://aiweekly.co/alerts/paper2agent-turns-74-of-100-biology-papers-into-ai-agents AI Weekly en 2026-09-24T09:12:01+00:00 Paper2Agent turns 74 of 100 biology papers into AI agents https://aiweekly.co/alerts/stanfords-paper2agent-turns-research-papers-into-mcp-agents AI Weekly en 2026-09-24T09:12:01+00:00 Stanford's Paper2Agent turns research papers into MCP agents https://aiweekly.co/alerts/paper2agent-converts-research-papers-into-interactive-ai-agents AI Weekly en 2026-09-24T09:12:01+00:00 Paper2Agent converts research papers into interactive AI agents https://aiweekly.co/alerts/paper2agent-turns-74-of-100-biology-papers-into-ai-agents AI Weekly en 2026-09-24T09:12:01+00:00 Paper2Agent turns 74 of 100 biology papers into AI agents https://aiweekly.co/alerts/stanfords-paper2agent-turns-papers-into-interactive-ai-agents AI Weekly en 2026-09-24T09:12:01+00:00 Stanford's Paper2Agent turns papers into interactive AI agents https://aiweekly.co/alerts/paper2agent-turns-research-papers-into-interactive-ai-agents AI Weekly en 2026-09-24T09:12:01+00:00 Paper2Agent turns research papers into interactive AI agents https://aiweekly.co/alerts/khlaaf-urges-nuclear-style-regulation-for-ai-developers AI Weekly en 2026-09-24T09:12:01+00:00 Khlaaf urges nuclear-style regulation for AI developers https://aiweekly.co/alerts/ai-nows-khlaaf-tells-nature-ai-firms-cant-self-regulate AI Weekly en 2026-09-24T09:12:01+00:00 AI Now's Khlaaf tells Nature: AI firms can't self-regulate https://aiweekly.co/alerts/stanfords-paper2agent-converts-74-of-100-papers-to-agents AI Weekly en 2026-09-24T09:12:01+00:00 Stanford's Paper2Agent converts 74 of 100 papers to agents https://aiweekly.co/alerts/nhtsa-probes-commaai-hands-off-tech-after-two-fatal-crashes AI Weekly en 2026-09-24T07:59:02+00:00 NHTSA Probes Comma.ai Hands-Off Tech After Two Fatal Crashes https://aiweekly.co/alerts/deepminds-kavukcuoglu-targets-pre-year-end-gemini-4-ship AI Weekly en 2026-09-24T06:38:20+00:00 DeepMind's Kavukcuoglu Targets Pre-Year-End Gemini 4 Ship https://aiweekly.co/alerts/nvidia-bar-ilan-post-reccar-to-close-video-attention-gap AI Weekly en 2026-09-24T06:38:20+00:00 NVIDIA, Bar-Ilan Post RecCAR to Close Video Attention Gap https://aiweekly.co/alerts/jitmem-curates-llm-agent-memory-at-read-time-not-write-time AI Weekly en 2026-09-24T06:38:20+00:00 JitMem Curates LLM Agent Memory at Read Time, Not Write Time https://aiweekly.co/alerts/studentbench-ai-tutors-match-human-gre-gains-at-918x-lower-cost AI Weekly en 2026-09-24T06:38:20+00:00 StudentBench: AI Tutors Match Human GRE Gains at 918x Lower Cost https://aiweekly.co/alerts/khlaaf-in-nature-ai-labs-cant-set-their-own-governance AI Weekly en 2026-09-24T06:12:01+00:00 Khlaaf in Nature: AI labs can't set their own governance https://aiweekly.co/alerts/ai-nows-khlaaf-calls-for-liability-not-ai-self-regulation AI Weekly en 2026-09-24T06:12:01+00:00 AI Now's Khlaaf calls for liability, not AI self-regulation https://aiweekly.co/alerts/openai-wires-chatgpt-voice-into-work-agent-and-gpt-6-models AI Weekly en 2026-09-24T06:12:01+00:00 OpenAI wires ChatGPT Voice into Work agent and GPT-6 models https://aiweekly.co/alerts/google-arda-eye-ai-passes-million-patients-across-three-countries AI Weekly en 2026-09-24T06:12:01+00:00 Google ARDA eye AI passes million patients across three countries https://aiweekly.co/alerts/openai-agents-probed-australian-health-data-transluce-says AI Weekly en 2026-09-24T06:12:01+00:00 OpenAI agents probed Australian health data, Transluce says https://aiweekly.co/alerts/paper-sycophancy-evals-conflate-receptiveness-with-deference AI Weekly en 2026-09-24T06:12:01+00:00 Paper: sycophancy evals conflate receptiveness with deference https://aiweekly.co/alerts/modal-in-talks-at-15b-baseten-at-26b-as-inference-booms AI Weekly en 2026-09-24T04:57:21+00:00 Modal in Talks at $15B, Baseten at $26B as Inference Booms https://aiweekly.co/alerts/membodied-beats-stateless-vla-policy-781x-on-rmbench-tasks AI Weekly en 2026-09-24T04:57:21+00:00 MemBodied Beats Stateless VLA Policy 7.81x on RMBench Tasks https://aiweekly.co/alerts/whatworkedbench-measures-ai-agents-experimental-understanding AI Weekly en 2026-09-24T04:57:21+00:00 WhatWorkedBench Measures AI Agents' Experimental Understanding https://aiweekly.co/alerts/schrodingerrepo-scrambles-repo-cues-swe-bench-scores-fall AI Weekly en 2026-09-24T03:53:35+00:00 SchrodingerRepo scrambles repo cues, SWE-bench scores fall https://aiweekly.co/alerts/chen-led-survey-maps-memory-mechanisms-in-ar-video-generation AI Weekly en 2026-09-24T03:26:47+00:00 Chen-Led Survey Maps Memory Mechanisms in AR Video Generation https://aiweekly.co/alerts/jill-lepore-legal-system-unprepared-for-autonomous-ai AI Weekly en 2026-09-24T03:12:01+00:00 Jill Lepore: Legal System Unprepared for Autonomous A.I. https://aiweekly.co/alerts/google-eye-ai-scales-from-one-hospital-to-a-million-patients AI Weekly en 2026-09-24T03:12:01+00:00 Google eye AI scales from one hospital to a million patients https://aiweekly.co/alerts/googles-diabetic-retinopathy-ai-passes-one-million-patients AI Weekly en 2026-09-24T03:12:01+00:00 Google's diabetic retinopathy AI passes one million patients https://aiweekly.co/alerts/adobe-amazon-workday-discount-ai-to-fend-off-anthropic-openai AI Weekly en 2026-09-24T03:12:01+00:00 Adobe, Amazon, Workday Discount AI to Fend Off Anthropic, OpenAI https://aiweekly.co/alerts/openai-agent-breached-medicare-data-portal-albanese-says AI Weekly en 2026-09-24T03:12:01+00:00 OpenAI agent breached Medicare data portal, Albanese says https://aiweekly.co/alerts/meta-muse-tops-chatgpts-early-ios-pace-hits-18m-downloads AI Weekly en 2026-09-24T03:12:01+00:00 Meta Muse tops ChatGPT's early iOS pace, hits 1.8M downloads https://aiweekly.co/alerts/john-warner-renews-case-against-ai-grading-student-writing AI Weekly en 2026-09-24T03:12:01+00:00 John Warner renews case against AI grading student writing https://aiweekly.co/alerts/grouptravelbench-tests-llm-agents-on-group-travel-plans AI Weekly en 2026-09-24T03:12:01+00:00 GroupTravelBench tests LLM agents on group travel plans https://aiweekly.co/alerts/nvidia-chip-emissions-match-russian-state-coal-producer-report-says AI Weekly en 2026-09-24T03:12:01+00:00 Nvidia chip emissions match Russian state coal producer, report says https://aiweekly.co/alerts/albanese-confronts-altman-after-openai-agent-hit-medicare-site AI Weekly en 2026-09-24T03:12:01+00:00 Albanese Confronts Altman After OpenAI Agent Hit Medicare Site https://aiweekly.co/alerts/skild-ais-s1-humanoid-learns-soccer-via-140-year-self-play AI Weekly en 2026-09-24T03:12:01+00:00 Skild AI's S1 humanoid learns soccer via 140-year self-play https://aiweekly.co/alerts/sanders-casar-bill-bans-superintelligence-adds-ai-department AI Weekly en 2026-09-24T02:00:57+00:00 Sanders, Casar Bill Bans Superintelligence, Adds AI Department https://aiweekly.co/alerts/openai-apples-chatgpt-in-siri-deal-persistently-underperformed AI Weekly en 2026-09-24T00:40:41+00:00 OpenAI: Apple's ChatGPT-in-Siri Deal 'Persistently Underperformed' https://aiweekly.co/alerts/alibaba-ships-qwen-audio-31-stack-cuts-voice-apis-up-to-95 AI Weekly en 2026-09-24T00:40:41+00:00 Alibaba Ships Qwen-Audio 3.1 Stack, Cuts Voice APIs Up to 95% https://aiweekly.co/alerts/black-forest-labs-open-weights-flux-3-action-a-7b-robotics-model AI Weekly en 2026-09-24T00:40:41+00:00 Black Forest Labs Open-Weights FLUX 3 Action, a 7B Robotics Model https://aiweekly.co/alerts/khlaaf-urges-liability-laws-for-negligent-ai-developers AI Weekly en 2026-09-24T00:12:01+00:00 Khlaaf urges liability laws for negligent AI developers https://aiweekly.co/alerts/biglam-packs-32b-token-europeana-newspapers-on-hugging-face AI Weekly en 2026-09-24T00:12:01+00:00 BigLAM packs 32B-token Europeana newspapers on Hugging Face https://aiweekly.co/alerts/openai-agent-accessed-australian-medicare-portal-pm-says AI Weekly en 2026-09-23T21:26:01+00:00 OpenAI Agent Accessed Australian Medicare Portal, PM Says https://aiweekly.co/alerts/dairs-alex-hanna-takes-on-five-ai-and-labor-myths AI Weekly en 2026-09-23T21:12:01+00:00 DAIR's Alex Hanna takes on five AI-and-labor myths https://aiweekly.co/alerts/ai-firms-cant-self-regulate-khlaaf-argues-in-nature AI Weekly en 2026-09-23T21:12:01+00:00 AI firms can't self-regulate, Khlaaf argues in Nature https://aiweekly.co/alerts/anthropic-exit-refuels-ai-extinction-debate-nature-reports AI Weekly en 2026-09-23T21:12:01+00:00 Anthropic exit refuels AI extinction debate, Nature reports https://aiweekly.co/alerts/anthropic-alignment-lead-puts-ai-extinction-odds-above-10 AI Weekly en 2026-09-23T21:12:01+00:00 Anthropic alignment lead puts AI extinction odds above 10% https://aiweekly.co/alerts/hawley-probes-flocks-120000-camera-ai-surveillance-network AI Weekly en 2026-09-23T21:12:01+00:00 Hawley probes Flock's 120,000-camera AI surveillance network https://aiweekly.co/alerts/doctorow-the-real-ai-hallucination-is-on-the-users-side AI Weekly en 2026-09-23T18:12:01+00:00 Doctorow: the real AI hallucination is on the user's side https://aiweekly.co/alerts/ai-now-scientist-labs-cant-be-trusted-to-self-regulate AI Weekly en 2026-09-23T18:12:01+00:00 AI Now scientist: labs can't be trusted to self-regulate https://aiweekly.co/alerts/nature-weighs-ai-extinction-warnings-after-anthropic-exit AI Weekly en 2026-09-23T18:12:01+00:00 Nature weighs AI extinction warnings after Anthropic exit https://aiweekly.co/alerts/google-retinal-ai-hits-1m-patients-across-three-countries AI Weekly en 2026-09-23T18:12:01+00:00 Google retinal AI hits 1M patients across three countries https://aiweekly.co/alerts/pilgrim-raises-25m-seed-at-150m-for-airborne-biothreat-sensor AI Weekly en 2026-09-23T16:56:40+00:00 Pilgrim raises $25M seed at $150M for airborne biothreat sensor https://aiweekly.co/alerts/paper-10-frontier-llms-collude-in-94-of-paired-agent-runs AI Weekly en 2026-09-23T15:54:07+00:00 Paper: 10 frontier LLMs collude in 94% of paired-agent runs https://aiweekly.co/alerts/scriptmoe-tops-vlm-baseline-on-229-language-ocr-benchmark AI Weekly en 2026-09-23T15:54:07+00:00 ScriptMoE tops VLM baseline on 229-language OCR benchmark https://aiweekly.co/alerts/hyperq-adds-quantum-circuits-to-11b-diffusion-lm-beats-lora AI Weekly en 2026-09-23T15:54:07+00:00 HyperQ Adds Quantum Circuits to 1.1B Diffusion LM, Beats LoRA https://aiweekly.co/alerts/jev-judge-model-matches-llm-accuracy-at-036-of-the-cost AI Weekly en 2026-09-23T15:31:38+00:00 JEV Judge Model Matches LLM Accuracy at 0.36% of the Cost https://aiweekly.co/alerts/sjtu-robofollow-exposes-vla-instruction-following-mirage AI Weekly en 2026-09-23T15:31:38+00:00 SJTU RoboFollow Exposes VLA Instruction-Following Mirage https://aiweekly.co/alerts/openais-hudson-valley-summer-camp-for-creators-backfires AI Weekly en 2026-09-23T15:12:01+00:00 OpenAI's Hudson Valley 'Summer Camp' for creators backfires https://aiweekly.co/alerts/khlaaf-put-ai-labs-under-nuclear-style-safety-regulators AI Weekly en 2026-09-23T15:12:01+00:00 Khlaaf: put AI labs under nuclear-style safety regulators https://aiweekly.co/alerts/ai-nows-khlaaf-urges-aviation-style-ai-oversight-in-nature AI Weekly en 2026-09-23T15:12:01+00:00 AI Now's Khlaaf urges aviation-style AI oversight in Nature https://aiweekly.co/alerts/nature-dissects-the-science-behind-ai-extinction-warnings AI Weekly en 2026-09-23T15:12:01+00:00 Nature dissects the science behind AI extinction warnings https://aiweekly.co/alerts/ema-raises-77m-series-b-says-agents-are-replacing-saas AI Weekly en 2026-09-23T13:56:44+00:00 Ema Raises $77M Series B, Says Agents Are Replacing SaaS https://aiweekly.co/alerts/carnegie-china-passes-us-as-top-ai-talent-hub-406-to-342 AI Weekly en 2026-09-23T13:56:44+00:00 Carnegie: China Passes US as Top AI Talent Hub, 40.6% to 34.2% https://aiweekly.co/alerts/uk-cma-wants-ai-assistants-on-googles-choice-screens AI Weekly en 2026-09-23T12:31:47+00:00 UK CMA Wants AI Assistants on Google's Choice Screens https://aiweekly.co/alerts/chris-hughes-proposes-federal-office-of-ai-supervision-in-ft AI Weekly en 2026-09-23T12:31:47+00:00 Chris Hughes Proposes Federal 'Office of AI Supervision' in FT https://aiweekly.co/alerts/bloomberg-china-ai-trade-rewards-export-facing-tech-36-vs-9 AI Weekly en 2026-09-23T12:31:47+00:00 Bloomberg: China AI Trade Rewards Export-Facing Tech, 36% vs 9% https://aiweekly.co/alerts/ai-nows-khlaaf-calls-for-nuclear-grade-ai-oversight AI Weekly en 2026-09-23T12:12:01+00:00 AI Now's Khlaaf calls for nuclear-grade AI oversight https://aiweekly.co/alerts/ai-nows-khlaaf-hold-ai-labs-liable-end-self-regulation AI Weekly en 2026-09-23T12:12:01+00:00 AI Now's Khlaaf: hold AI labs liable, end self-regulation https://aiweekly.co/alerts/amie-team-benchmarks-cant-earn-clinical-ai-trust AI Weekly en 2026-09-23T12:12:01+00:00 AMIE team: benchmarks can't earn clinical AI trust https://aiweekly.co/alerts/on-prem-clinical-ai-agent-hits-90-on-mimic-iv-diagnosis-test AI Weekly en 2026-09-23T12:12:01+00:00 On-prem clinical AI agent hits 90% on MIMIC-IV diagnosis test https://aiweekly.co/alerts/amie-team-benchmarks-alone-cant-earn-clinical-ais-trust AI Weekly en 2026-09-23T12:12:01+00:00 AMIE team: benchmarks alone can't earn clinical AI's trust https://aiweekly.co/alerts/nature-paper-on-prem-clinical-ai-matches-gpt-52-diagnosis AI Weekly en 2026-09-23T12:12:01+00:00 Nature paper: on-prem clinical AI matches GPT-5.2 diagnosis https://aiweekly.co/alerts/googles-amie-team-clinical-ai-trust-cant-be-benchmarked-in AI Weekly en 2026-09-23T12:12:01+00:00 Google's AMIE team: clinical AI trust can't be benchmarked in https://aiweekly.co/alerts/on-prem-clinical-ai-matches-cloud-gpt-52-in-nature-medicine AI Weekly en 2026-09-23T12:12:01+00:00 On-prem clinical AI matches cloud GPT-5.2 in Nature Medicine https://aiweekly.co/alerts/qwen-35-on-prem-agent-matches-gpt-52-on-diagnostic-test AI Weekly en 2026-09-23T12:12:01+00:00 Qwen-3.5 on-prem agent matches GPT-5.2 on diagnostic test https://aiweekly.co/alerts/on-premise-clinical-ai-agents-match-gpt-52-within-07-points AI Weekly en 2026-09-23T12:12:01+00:00 On-premise clinical AI agents match GPT-5.2 within 0.7 points https://aiweekly.co/alerts/amie-team-benchmarks-alone-cant-earn-medical-ais-trust AI Weekly en 2026-09-23T12:12:01+00:00 AMIE team: benchmarks alone can't earn medical AI's trust https://aiweekly.co/alerts/on-premise-qwen-35-matches-cloud-gpt-52-on-ed-diagnosis AI Weekly en 2026-09-23T12:12:01+00:00 On-premise Qwen-3.5 matches cloud GPT-5.2 on ED diagnosis https://aiweekly.co/alerts/google-amie-team-only-prospective-trials-earn-clinical-ai-trust AI Weekly en 2026-09-23T12:12:01+00:00 Google AMIE team: only prospective trials earn clinical AI trust https://aiweekly.co/alerts/nature-examines-the-science-behind-ai-extinction-claims AI Weekly en 2026-09-23T12:12:01+00:00 Nature examines the science behind AI extinction claims https://aiweekly.co/alerts/tekever-closes-580m-series-d-at-64b-eyes-defense-ma AI Weekly en 2026-09-23T10:58:12+00:00 Tekever closes $580M Series D at $6.4B, eyes defense M&A https://aiweekly.co/alerts/metas-muse-triggers-selloff-in-consumer-inertia-stocks AI Weekly en 2026-09-23T09:33:31+00:00 Meta's Muse Triggers Selloff in 'Consumer Inertia' Stocks https://aiweekly.co/alerts/cansemi-files-for-919m-chinext-ipo-amid-losses-at-guangzhou-fab AI Weekly en 2026-09-23T09:33:31+00:00 CanSemi Files for $919M ChiNext IPO Amid Losses at Guangzhou Fab https://aiweekly.co/alerts/paper2agent-converts-74-biology-papers-into-working-ai-agents AI Weekly en 2026-09-23T09:12:01+00:00 Paper2Agent converts 74 biology papers into working AI agents https://aiweekly.co/alerts/nature-paper-paper2agent-turns-research-articles-into-ai-agents AI Weekly en 2026-09-23T09:12:01+00:00 Nature paper: Paper2Agent turns research articles into AI agents https://aiweekly.co/alerts/paper2agent-turns-research-papers-into-interactive-ai-agents AI Weekly en 2026-09-23T09:12:01+00:00 Paper2Agent turns research papers into interactive AI agents https://aiweekly.co/alerts/paper2agent-converts-research-papers-into-interactive-ai-agents AI Weekly en 2026-09-23T09:12:01+00:00 Paper2Agent converts research papers into interactive AI agents https://aiweekly.co/alerts/paper2agent-turns-research-papers-into-interactive-ai-agents AI Weekly en 2026-09-23T09:12:01+00:00 Paper2Agent turns research papers into interactive AI agents https://aiweekly.co/alerts/mathematicians-use-ai-to-solve-last-open-m23-galois-case AI Weekly en 2026-09-23T08:07:26+00:00 Mathematicians Use AI to Solve Last Open M23 Galois Case https://aiweekly.co/alerts/waymo-refunds-285-for-bay-area-transit-plus-robotaxi-trips AI Weekly en 2026-09-23T06:31:29+00:00 Waymo Refunds $2.85 for Bay Area Transit-Plus-Robotaxi Trips https://aiweekly.co/alerts/stablevq-reports-100-codebook-use-rfid-105-on-imagenet AI Weekly en 2026-09-23T06:31:29+00:00 StableVQ Reports 100% Codebook Use, rFID 1.05 on ImageNet https://aiweekly.co/alerts/trump-admin-calls-australia-duty-of-care-bill-censorship AI Weekly en 2026-09-23T06:31:29+00:00 Trump Admin Calls Australia Duty of Care Bill 'Censorship' https://aiweekly.co/alerts/taste-bench-puts-best-llm-agent-at-597-on-decision-forks AI Weekly en 2026-09-23T06:31:29+00:00 Taste-Bench Puts Best LLM Agent at 59.7% on Decision Forks https://aiweekly.co/alerts/typesafe-ais-jev-outputs-numbers-not-text-at-0042m-input AI Weekly en 2026-09-23T06:12:01+00:00 TypeSafe AI's Jev outputs numbers, not text, at $0.042/M input https://aiweekly.co/alerts/openai-anthropic-slash-prices-as-gpt-6-and-opus-55-ship AI Weekly en 2026-09-23T06:12:01+00:00 OpenAI, Anthropic slash prices as GPT-6 and Opus 5.5 ship https://aiweekly.co/alerts/devezer-buzbas-replicability-itself-isnt-replicable AI Weekly en 2026-09-23T06:12:01+00:00 Devezer & Buzbas: replicability itself isn't replicable https://aiweekly.co/alerts/alibaba-cloud-to-open-first-regions-in-turkey-finland-netherlands AI Weekly en 2026-09-23T04:56:39+00:00 Alibaba Cloud to Open First Regions in Turkey, Finland, Netherlands https://aiweekly.co/alerts/aide-agent-self-rewrites-cut-reward-hacking-to-32-in-8-days AI Weekly en 2026-09-23T04:56:39+00:00 AIDE² Agent Self-Rewrites Cut Reward Hacking to 32% in 8 Days https://aiweekly.co/alerts/ant-groups-realtime-venus-9b-tops-six-of-eight-video-tests AI Weekly en 2026-09-23T03:54:08+00:00 Ant Group's Realtime-Venus 9B tops six of eight video tests https://aiweekly.co/alerts/flash-dllm-posts-11x-humaneval-51x-gsm8k-dllm-speedups AI Weekly en 2026-09-23T03:54:08+00:00 Flash-dLLM posts 11x HumanEval, 5.1x GSM8K dLLM speedups https://aiweekly.co/alerts/tencent-arcs-gae-halves-3d-camera-error-in-world-generation AI Weekly en 2026-09-23T03:54:08+00:00 Tencent ARC's GAE halves 3D camera error in world generation https://aiweekly.co/alerts/ruler-matches-deepseek-v3-on-svg-using-text-only-rubrics AI Weekly en 2026-09-23T03:54:08+00:00 RULER matches DeepSeek-V3 on SVG using text-only rubrics https://aiweekly.co/alerts/ovis-embedding-one-shared-backbone-tops-five-omni-modal-suites AI Weekly en 2026-09-23T03:54:08+00:00 Ovis-Embedding: one shared backbone tops five omni-modal suites https://aiweekly.co/alerts/bank-of-america-capital-one-flag-ai-shopping-agent-risks AI Weekly en 2026-09-23T03:12:01+00:00 Bank of America, Capital One Flag AI Shopping Agent Risks https://aiweekly.co/alerts/stellar-colosseum-agent-hits-71-on-tcs-bench-218222-codeforces AI Weekly en 2026-09-23T00:12:01+00:00 Stellar Colosseum agent hits 71% on TCS-Bench, 218/222 Codeforces https://aiweekly.co/alerts/chronologic-benchmarks-llms-on-1831-1930-english-contexts AI Weekly en 2026-09-23T00:12:01+00:00 Chronologic benchmarks LLMs on 1831-1930 English contexts https://aiweekly.co/alerts/underwood-lab-releases-866-question-1831-1930-llm-benchmark AI Weekly en 2026-09-23T00:12:01+00:00 Underwood lab releases 866-question 1831-1930 LLM benchmark https://aiweekly.co/alerts/bessent-emerges-as-frontrunner-for-trumps-ai-czar-role AI Weekly en 2026-09-22T22:56:34+00:00 Bessent Emerges as Frontrunner for Trump's AI Czar Role https://aiweekly.co/alerts/tiktok-usds-joins-lantern-pledges-2m-to-cybertipline-overhaul AI Weekly en 2026-09-22T21:27:18+00:00 TikTok USDS Joins Lantern, Pledges $2M to CyberTipline Overhaul https://aiweekly.co/alerts/gebru-bender-ais-summer-breakthroughs-fall-to-scrutiny AI Weekly en 2026-09-22T21:27:18+00:00 Gebru, Bender: AI's Summer 'Breakthroughs' Fall to Scrutiny https://aiweekly.co/alerts/jev-decision-index-ranks-30-open-weight-decision-models AI Weekly en 2026-09-22T21:12:01+00:00 Jev Decision Index ranks 30+ open-weight decision models https://aiweekly.co/alerts/pentagon-probe-cites-palantir-maven-in-iran-school-strike AI Weekly en 2026-09-22T21:12:01+00:00 Pentagon probe cites Palantir Maven in Iran school strike https://aiweekly.co/alerts/amd-ships-14-zebra-hylo-checkpoints-cuts-kv-cache-92-98 AI Weekly en 2026-09-22T19:56:59+00:00 AMD Ships 14 Zebra-HyLo Checkpoints, Cuts KV-Cache 92-98% https://aiweekly.co/alerts/aclarena-diagnoses-non-monotonic-forgetting-in-agent-post-training AI Weekly en 2026-09-22T19:56:59+00:00 ACLArena Diagnoses Non-Monotonic Forgetting in Agent Post-Training https://aiweekly.co/alerts/openai-opens-model-training-to-outside-safety-evaluators AI Weekly en 2026-09-22T18:32:27+00:00 OpenAI Opens Model Training to Outside Safety Evaluators https://aiweekly.co/alerts/data-society-60-of-americans-want-to-cap-new-data-centers AI Weekly en 2026-09-22T18:12:01+00:00 Data & Society: 60% of Americans want to cap new data centers https://aiweekly.co/alerts/ai-graders-penalize-ai-generated-prose-25x-more-than-humans AI Weekly en 2026-09-22T18:12:01+00:00 AI graders penalize 'AI-generated' prose 2.5x more than humans https://aiweekly.co/alerts/fei-urges-trump-xi-summit-to-address-ai-enabled-worms AI Weekly en 2026-09-22T18:12:01+00:00 Fei urges Trump-Xi summit to address AI-enabled worms https://aiweekly.co/alerts/anthropic-ships-claude-opus-55-40-cheaper-than-opus-5 AI Weekly en 2026-09-22T18:12:01+00:00 Anthropic ships Claude Opus 5.5, 40% cheaper than Opus 5 https://aiweekly.co/alerts/talos-releases-cairn-uncovers-llm-voting-malware-closedquorum AI Weekly en 2026-09-22T18:12:01+00:00 Talos releases CAIRN, uncovers LLM-voting malware CLOSEDQUORUM https://aiweekly.co/alerts/arenaflow-proposes-hierarchical-credit-for-open-ended-agent-rl AI Weekly en 2026-09-22T18:12:01+00:00 ArenaFlow proposes hierarchical credit for open-ended agent RL